Loughborough University
Thesis.pdf (1.7 MB)

Rethinking construction safety

Download (1.7 MB)
posted on 2019-06-03, 15:26 authored by Eleanor Harvey
Recent human-induced disasters have prompted concerns that conventional risk management (quantifying and controlling hazards) is not only reaching its limits but introducing vulnerabilities which increase the probability of accidents. These compliance or behaviour-based approaches can obscure risks, suppress data, and detract from major hazards — feeding complacency and reducing resilience as the workforce becomes disfranchised and deskilled. In response to this, emerging constructs favour non-prescriptive methods for managing risk and look to support the human factors which underpin these. This is known as the ‘fifth’ or ‘Adaptive’ age of safety.
Adaptive safety differs from its predecessors in two fundamental ways: Firstly, it refutes notions of accident causation as predictable and, consequently, it casts doubt on our understanding of the human contribution to accidents. Both these undermine traditional risk management models. Instead, an approach is advocated which accounts for complex systemic interactions and interpersonal aspects of risk management. As yet there is a lack of evidence and guidance to support Adaptive safety in industry.
The construction sector is particularly challenging with respect to safety because of its litigation culture, affinity for traditional safety, project-based network structure, and transient workforce. The emerging Adaptive paradigm could potentially be a valuable opportunity to improve construction safety: Moving towards engaging workers, drawing on their expertise, increasing vigilance, and enabling a response to the unexpected; however, its applicability to this unique sector is disputed.
This research uses mixed qualitative methods to examine the compatibility of Adaptive safety with construction, theoretically and then in practice. The first part of this thesis explores leadership as a means to foster an Adaptive culture in organisations. A framework for ‘Adaptive safety leadership’ is synthesised from literature on safety leadership, Safety Intelligence, and Adaptive safety, and evaluated against construction practice in an interview study. The findings show many safety practitioners and policy-makers in construction recognise the need for systems thinking and relational aspects of leadership, but the structure, pressure, and culture of the industry mean the tendency to blame workers and bureaucratise risk is difficult to overcome. The relationships between these systemic ‘Originating Influences’ and the ‘Immediate Circumstances’ surrounding accidents are crystallised in an update of Haslam, Hide, Gibb, Gyi, Pavitt, Atkinson and Duff’s (2005) contributing factors in construction accidents (ConCA) model — explaining the challenges to taking an Adaptive approach and demonstrating the value of systems thinking in construction.
The second part of this thesis case studies Adaptive safety in practice, following two pioneering infrastructure megaprojects as they embrace Dekker’s (2017a) ‘Safety Differently’. Proactivity, relationships, communication, and job-satisfaction are improving as the new philosophy is cultivated and allowed to evolve in collaboration with the workforce. The role of safety leaders in embedding this concept in frontline work and factors contributing to their success are examined. Aspects of the company, project, and London’s megaproject ecology have meant Laing O’Rourke has been uniquely well-equipped to make this transition, but the sector’s pace of change and persistent culture have been challenging.
These insights contribute an improved understanding of the mechanisms of Adaptive safety and the factors which support and hinder its success with a view to its wider implementation. However, the work also warns against substituting engagement for safety; questions the ethics of responsibilising workers; and stresses the need for a context-sensitive balance of new and old safety paradigms. It highlights the inadequacies of this construct which need to be resolved before it can be operationalised.



  • Design


Loughborough University

Rights holder

© Eleanor J. Harvey

Publication date



A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of the degree of Doctor of Philosophy at Loughborough University.


  • en


Patrick Waterson ; Andrew Dainty

Qualification name

  • PhD

Qualification level

  • Doctoral

This submission includes a signed certificate in addition to the thesis file(s)

  • I have submitted a signed certificate